NorCom extends DaSense with Deep Learning Module

The DaSense Deep Learning module has been installed at a German OEM since the beginning of the month and is used in the field of autonomous driving. The application trains neural networks for autonomous driving at various locations worldwide.

Deep learning in five steps

The DaSense Deep Learning module packs the elaborate deep learning workflow into a few, well-defined steps. This results in a simple and fast workflow without any breaks.

First, a new neural network is developed locally for the specific application and tested with only a small amount of data for correct processability. Together with the necessary training parameters, an experiment is created from this and packed into a Deep Learning App. The Deep Learning app sends the training code to all available clusters – worldwide if needed. This enables a parallel execution of training – locally or in the cloud. Finally, the results are collected, evaluated and the model improved accordingly. A new training cycle can begin.The user can keep track of the status and execution times of deep learning jobs, use of resources, and intelligent workload distribution online at any time and intervene at any time.

"The DaSense Deep Learning module enables a very powerful and fast iteration and improvement cycle, as the training of models as distributed jobs is performed in parallel on all available clusters," explains Dr. Tobias Abthoff, Managing Director of NorCom. "Especially with a large project, DaSense can fully exploit its advantages: the fast processing of large amounts of data, regardless of location. The deep learning theme is also one of the most exciting things the market has to offer – and we look forward to working with a reputable company. "

Independent use possible

The Deep Learning module is not necessarily tied to a DaSense installation but can also be used as a stand-alone application. It integrates big data technologies with proven NorCom technology for scalability and security when processing very large amounts of data, enabling the successful training of neural networks. In an existing DaSense installation, the deep learning module can be seamlessly integrated.

About DaSense

NorCom’s DaSense technology analyzes and manages complex, heterogeneous, semi-structured, and highly proprietary data, specializing in technical features and challenges in the automotive industry. So far, the software has been used primarily in the area of ??research and development. DaSense’s scope of application covers the entire range of automakers‘ current and future products: machining, management and analysis of data can be performed on classic drive concepts, from hybrid and electric cars to autonomous vehicles through DaSense.